天津农学院学报 ›› 2014, Vol. 21 ›› Issue (2): 33-38.

• 研究与简报 • 上一篇    下一篇

基于灰度共生矩阵的纹理分析的统计方法

陈成钢1, 艾涛2   

  1. 1. 天津城建大学 理学院,天津 300384;
    2. 上海中兴通讯技术有限责任公司,上海 201203
  • 收稿日期:2013-01-28 发布日期:2019-10-21
  • 作者简介:陈成钢(1976-),男,山东菏泽人,副教授,硕士,主要从事概率统计方向研究。E-mail:chenchenggang@yeah.net。

Statistical Method of Texture Analysis Based on Gray Co-occurrence Matrix

CHEN Cheng-gang1, AI Tao2   

  1. 1. College of Fundamental Subject, TIUC, Tianjin 300384, China;
    2. Shanghai Zhongxing Communication Technology Limited Liability Company, Shanghai 201203, China
  • Received:2013-01-28 Published:2019-10-21

摘要: 由于影像图像纹理性强,这使得对于压缩后图像纹理质量的要求显得更加严格。然而,目前的影像图像压缩技术对压缩后图像的纹理效果处理不佳。本文将图像的灰度信息加进灰度共生矩阵,使得灰度矩阵能包含图像的纹理基元及其排列信息,并给出了统计分析方法。结果表明,与传统的压缩方法相比,文中提出的新方法更好地保留了原始图像的纹理效果及其相关信息。

关键词: 影像图像, 灰度共生矩阵, 纹理分析, 统计方法

Abstract: The strong texture characteristics of image produce more stringent requirements for the texture qualities of the compressed image. However, the treatment effect of the compressed image texture is poor based on the current video compression technology. The experiment puts the gray information of images into gray co-occurrence matrix and the gray matrix can contain the image texton and arrangement information, and the method of statistical analysis is given. The results show that compared with the traditional compression methods, the new method proposed in this paper is better in retaining the texture effect and relevant information of original images.

Key words: image, gray co-ocurrence matrix, texture analysis, statistical method

中图分类号: